Commit Graph

24 Commits

Author SHA1 Message Date
Francesco Gatti f6527f51e3 Add csresnext50-panet-spp test. Works on CUDNN. Does not work with TensorRT
Signed-off-by: Francesco Gatti <gattifrancesco@hotmail.it>
2020-03-11 18:37:26 +01:00
xavier 38a1b9dcb2 Add Mobilenet2SSDLite test
The new test works both with TensorRT and cuDNN. Preprocessing and
Postprocessing are missing. Add ClippedReLU (for ReLU6), groups for
Conv2d, additional bias for convolution.

Other minors:
-move the timer in the detector to measure all the
processing time for a given frame (both centernet and yolo);
-add int8 flag.

Signed-off-by: Micaela Verucchi <micaelaverucchi@gmail.com>
Davide Sapienza <sapienza.dav@gmail.com>
2020-02-21 10:45:46 +01:00
Davide Sapienza 97b88ef52d Fix a bug in the deformable kernel.
There was a wrong variable initialization.

Fixes: 51ffcb1f50 ("Optimize deformable kernel")

Signed-off-by: Davide Sapienza <sapienza.dav@gmail.com>
2020-02-11 16:28:17 +01:00
Davide Sapienza 51ffcb1f50 Optimize deformable kernel
Signed-off.by: Ignacio Sañudo Olmedo<ignacio.sanudoolmedo@unimore.it>
Signed-off-by: Davide Sapienza <sapienza.dav@gmail.com>
2020-02-11 10:06:08 +01:00
Davide Sapienza 2503eba173 Fix memory leak
This commit moves cublasCreate out from dcn_v2_cuda_forward
to save some milliseconds and it adds cublasDestroy (cause
of memory leak).

Signed-off-by: Davide Sapienza <sapienza.dav@gmail.com>
2020-02-06 23:09:12 +01:00
Davide Sapienza 7f239efdc0 Centernet: fix pooling problem, add centrnet demo
Signed-off-by: Micaela Verucchi <micaelaverucchi@gmail.com>
Signed-off-by: Davide Sapienza <sapienza.dav@gmail.com>
2020-01-21 12:50:18 +01:00
fbagni 7e7f480e2b Yolov3_tiny works on tensorRT
Signed-off-by: fbagni <gattinomicino>
2019-12-23 16:11:41 +01:00
nvidia f0ea2c027f Merge branch 'cnet' of https://github.com/ceccocats/tkDNN into cnet 2019-12-20 15:27:55 +01:00
nvidia 854a4c316a Add ResizeLayerRT plugin
Signed-off-by: nvidia <micaelaverucchi@gmail.com>
2019-12-20 15:27:41 +01:00
Davide Sapienza e99b353d8b Fix the inference operation of the deformable convolutional layer.
This commit removes the malloc operation in the inference
method and adds the sigmoid kernel.

Signed-off-by: Davide Sapienza <sapienza.dav@gmail.com>
2019-12-19 14:47:57 +01:00
Davide Sapienza df888a3457 Add CenterNet based on Resnet101, TensorRT not implemented.
Signed-off-by: Davide Sapienza <sapienza.dav@gmail.com>
2019-12-18 18:59:54 +01:00
Micaela Verucchi 8b64d1876f deformable conv cu 2019-10-30 10:20:04 +01:00
Francesco Gatti c8f2e1b448 upsample ok 2018-12-20 17:08:31 +01:00
Francesco Gatti e91db28756 shortcut cu 2018-12-20 09:53:17 +01:00
Francesco Gatti 2cf8d8f6fc fp16 implementation, TODO deallocate in LayerWgs 2017-08-30 14:37:25 +00:00
Francesco Gatti 3124f86878 stream in TRT plugin 2017-08-10 19:21:15 +02:00
Francesco Gatti 9a6058ac4a removed sync 2017-08-10 18:47:21 +02:00
Francesco Gatti 266330009c opencv viz 2017-08-10 16:22:17 +02:00
Francesco Gatti ed5e5d58b5 TensorRT version 2017-08-01 17:51:49 +02:00
Francesco Gatti b94931f9f7 yolo layers 2017-08-01 16:08:56 +02:00
Francesco Gatti 0a0acb9b1d test ready 2017-06-29 12:26:51 +00:00
Francesco Gatti b9f894ca75 better elu kernel 2017-06-28 15:16:38 +00:00
Francesco Gatti 3cb126420c conv2d implementation 2017-06-28 14:04:43 +00:00
Francesco Gatti a86df107be ELU implemented 2017-06-28 11:56:43 +00:00